Cheng et al. BMC Cancer (2017) 17:164 DOI 10.1186/s12885-017-3152-5

RESEARCHARTICLE Open Access A novel FOXO1-mediated dedifferentiation blocking role for DKK3 in adrenocortical carcinogenesis Joyce Y. Cheng1†, Taylor C. Brown1†, Timothy D. Murtha1, Adam Stenman3, C. Christofer Juhlin3, Catharina Larsson3, James M. Healy1, Manju L. Prasad2, Wolfram T. Knoefel4, Andreas Krieg4, Ute I. Scholl5, Reju Korah1 and Tobias Carling1,6*

Abstract Background: Dysregulated WNT signaling dominates adrenocortical malignancies. This study investigates whether silencing of the WNT negative regulator DKK3 (-related 3), an implicated adrenocortical differentiation marker and an established tumor suppressor in multiple cancers, allows dedifferentiation of the adrenal cortex. Methods: We analyzed the expression and regulation of DKK3 in human adrenocortical carcinoma (ACC) by qRT-PCR, immunofluorescence, promoter methylation assay, and copy number analysis. We also conducted functional studies on ACC cell lines, NCI-H295R and SW-13, using siRNAs and enforced DKK3 expression to test DKK3’sroleinblocking dedifferentiation of adrenal cortex. Results: While robust expression was observed in normal adrenal cortex, DKK3 was down-regulated in the majority (>75%) of adrenocortical carcinomas (ACC) tested. Both genetic ( copy loss) and epigenetic (promoter methylation) events were found to play significant roles in DKK3 down-regulation in ACCs. While NCI-H295R cells harboring β-catenin activating mutations failed to respond to DKK3 silencing, SW-13 cells showed increased motility and reduced clonal growth. Conversely, exogenously added DKK3 also increased motility of SW-13 cells without influencing their growth. Enforced over-expression of DKK3 in SW-13 cells resulted in slower cell growth by an extension of G1 phase, promoted survival of microcolonies, and resulted in significant impairment of migratory and invasive behaviors, largely attributable to modified cell adhesions and adhesion kinetics. DKK3-over-expressing cells also showed increased expression of Forkhead Box Protein O1 (FOXO1) transcription factor, RNAi silencing of which partially restored the migratory proficiency of cells without interfering with their viability. Conclusions: DKK3 suppression observed in ACCs and the effects of manipulation of DKK3 expression in ACC cell lines suggest a FOXO1-mediated differentiation-promoting role for DKK3 in the adrenal cortex, silencing of which may allow adrenocortical dedifferentiation and malignancy. Keywords: DKK3,FOXO1,Adrenocortical carcinogenesis

* Correspondence: [email protected] †Equal contributors 1Department of Surgery & Yale Endocrine Neoplasia Laboratory, Yale University School of Medicine, New Haven, CT, USA 6Department of Surgery, Yale University School of Medicine, 333 Cedar Street, FMB130A, New Haven, CT 06520, USA Full list of author information is available at the end of the article

© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Cheng et al. BMC Cancer (2017) 17:164 Page 2 of 13

Background suggesting a global tumor suppressor role for this WNT Adrenocortical carcinoma (ACC) is a rare (0.5–2 cases regulator (reviewed in [26]). Furthermore, ectopic per million/year) endocrine malignancy that carries a expression of DKK3 in a variety of types stifled poor prognosis at diagnosis due to its propensity to aggressive malignant behavior, reversed epithelial- metastasize before detection. Even with aggressive surgi- mesenchymal transition (EMT), and impaired cell motility, cal and oncologic therapy, the 5-year survival rate is an pointing towards a comprehensive dedifferentiation- abysmal 16–38% [1–4]. A major reason for the lack of blocking role for DKK3 [28, 29]. This study investigates a an effective targeted treatment strategy for ACCs is an potential tumor suppressor role for the implicated adrenal inadequate understanding of the molecular pathogenesis differentiation factor DKK3 in blocking dedifferentiation of of the disease [3, 4]. adrenocortical cells. Genetic and epigenetic dysregulations of the WNT, p53, and IGF2 pathways appear to dominate various cancer- Methods driving anomalies in the majority of ACCs [5–7]. Recent Tissue acquisition findings from comprehensive genetic analyses of ACCs Written informed consent was obtained from patients confirmed a signature role for WNT dysregulation in the prior to surgical resection of adrenal tissue according to origin and/or progression of ACCs [4, 6, 8, 9]. Physiologic- protocols approved by Institutional Review Boards at (a) ally, both canonical and non-canonical WNT signaling Yale University, New Haven, CT, USA, (b) Heinrich Heine pathways play global and zone-specific roles in the devel- University Düsseldorf, Düsseldorf, Germany, and (c) Karo- opment, differentiation, and homeostasis of the adrenal linska Institutet, Stockholm, Sweden. Tissue samples were gland [10, 11]. In particular, endocrine homeostasis of the flash-frozen (FF) in liquid nitrogen and stored at −80 °C adrenal glomerulosa and fasciculata zones is largely until processed for study. Specimens displaying unequivo- controlled by WNT-differentiation signaling mediated by cal histopathological characteristics of ACCs (n = 38) and the Wnt4-Fz1/2-Dvl3-β-Catenin-SF1 axis [12–16]. Regula- histologically normal adrenal tissue (n = 14) samples ex- tory components of this proposed adrenal cortex-specific cised with ACAs were selected for study. Consecutive un- Wnt4 axis include the secretory factors, frizzled-related stained/hematoxylin & eosin (H&E) stained 5 μM sections protein 1 (SFRP1) and the putative tumor suppressor, of formalin-fixed, paraffin-embedded (FFPE) tissue sam- DKK3 [14, 17, 18]. Aberrant WNT signaling has been ples underwent immunohistochemistry analyses. All sam- well-established in the origin of many tumor types and is ples were histopathologically confirmed by experienced strongly associated with stabilization of β-catenin in the endocrine pathologists before processing. cytoplasm and/or in the nucleus and constitutive activa- tion of WNT target [19, 20]. Similar stabilization DNA, RNA, and protein preparation and nuclear accumulation of β-catenin is seen in benign ad- Genomic DNA and total RNA were isolated from FF sam- renocortical adenomas (ACAs) and frequently in malignant ples using AllPrep DNA/RNA/Protein Mini Kit (Qiagen) ACCs [10, 21]. However, only 10% of ACCs with constitu- as per manufacturer’s recommendations. Quantity and tively active β-catenin carry mutations in the β-catenin gene quality of prepared nucleic acids were assessed by spectro- (CTNNB1), suggesting alternate mechanisms of aberrant photometry (NanoDrop Technologies, Inc.). Total protein WNT activation, including dysregulation of WNT inhibi- from cultured cells was extracted using Laemmli buffer tors such as Wif-1 [22]. Other WNT regulatory mutations (BioRad) as cell lysis buffer; protein concentrations were found in ACCs include PRKAR1A [23] and recently identi- quantified using Pierce BCA Protein Assay Kit (Thermo- fied KREMEN1 and ZNRF3 gene deletions [8, 24]. Fisher Scientific) and GloMax multidetection system (Pro- Although implicated in zonal differentiation and hor- mega), as per manufacturer’s instructions. mone biosynthesis [14, 25], a definitive role for the ubiqui- tous WNT inhibitor DKK3 in promoting functional Gene expression analysis differentiation and/or blocking tumor dedifferentiation of Total RNA (100 ng) was reverse transcribed using iScript the adrenal cortex has yet to be clarified. The inhibitory cDNA synthesis kit (Bio-Rad) as per manufacturer’sin- role of DKK3 in WNT signaling is context-dependent and structions. Quantitative real-time PCR (qRT-PCR) was per- appears to be influenced by a repertoire of cell surface re- formed in triplicate using TaqMan PCR master mix with ceptors and co-expressed ligands [26]. DKK3, a 38 kDa se- FAM fluorophore and probe/primer pairs specific to hu- creted glycoprotein with an N-terminal signal peptide, is man DKK3 (Hs00951307_m1), FOXO1 (Hs01054576_m1), also implicated in eliciting distinct intracellular roles in and RPLP0 (Hs99999902_m1) (ThermoFisher Scientific) addition to its secretory functions [27]. Reduced DKK3 ex- according to manufacturer’s cycling conditions using pression is observed in a variety of solid tumors, and re- CFX96 thermal cyclers (Bio-Rad). Gene expression levels expression studies in multiple cancer cell types mostly re- were normalized to mean RPLP0 expression levels. Relative sulted in cell cycle arrest and/or apoptosis, strongly gene expression values were calculated using recommended Cheng et al. BMC Cancer (2017) 17:164 Page 3 of 13

Livak method (Bio-Rad). Fold-change expression values antibodies and anti-goat FITC (fluorescein isothiocyan- were calculated by base-two logarithmic transformations of ate) and anti-mouse TR (Texas Red) secondary anti- relative gene expression values. bodies (1:1000) were used, followed by Ultracruz For pathway-focused gene expression analysis, (a) RT2 mounting agent containing 4′,6-diamidino-2-phenylin- Profile PCR Array Human WNT Signaling Pathway and dole (DAPI) (all from Santa Cruz Biotechnology, Inc.) (b) RT2 Profiler PCR Array Human Transcription Factors for indirect immunodetection. A Zeiss AX10 confocal were used according to protocol outlined in RT2 Profiler microscope with AxioVision 4.8 program was used for PCR Array Handbook (Qiagen). Briefly, 100 ng of DNA- IF analysis, and photomicrographs were taken at a total free RNA from each sample was used for 84 target genes magnification of 100× or 400×, as noted. listed in gene lists (available at www.qiagen.com) using 96-well RT2 profiler array format D. cDNA was prepared Cell culture, expression vectors, transfections, and using RT2 first strand kit and amplified using RT2 SYBR western blot detection Green Mastermix (both from Qiagen) using CFX96 ther- American Type Culture Collection (ATCC)-authenticated mal cycler. Differential expression of target genes was cal- human ACC cell lines SW-13 (CCL-105) and NCI-H295R culated using ΔΔCT method on data web portal at (CRL-2128) were maintained in growth conditions recom- www.SABiosciences.com/pcrarraydataanalysis.php. mended by ATCC, as reported previously [31]. For DKK3 treatments, a working concentration of 5 μg/mL (in PBS) Methylation-specific PCR of human recombinant DKK3 (R&D Systems) was used. Methylation status of CpG island A of DKK3 promoter RNAi silencing was carried out with 3 unique 27-mer (Chr11:12029737–12030841) was assessed by MethylScreen siRNA duplexes (designated siA, siB, and siC) targeting technology using EpiTect Methyl II PCR Assay (Qiagen) as DKK3 (Human) and FOXO1 (Human) transcripts. Univer- previously described [30]. Briefly, 125 ng of genomic DNA sal scrambled negative control siRNA was used as non- was mock-digested or digested with methylation-sensitive specific control (all from Origene). Lipofectamine2000- and methylation-dependent restriction enzymes individu- mediated transfection was carried out in Opti-MEM ac- ally or together, and methylation status of target DNA se- cording to manufacturer’s recommendations (Thermo- quence was measured using qRT-PCR with probes specific Fisher Scientific) in 6-well plates with starting densities of to target DKK3 promoter sequence. CT values were con- 50,000 cells/well for SW-13 and 80,000 cells/well for NCI- verted into percentages of unmethylated, intermediate- H295R. Transfection medium was replaced with regular methylated, and hypermethylated CpG values using a quan- growth medium after 24 h of transfection. Cells were lysed titation algorithm from EpiTect Methyl II PCR Assay for RNA extraction (after 24 h) or protein extraction (after Handbook (Qiagen). Tissue samples were designated as 48 h), and assays were done 48 h post-transfection. hypermethylated (>5% alleles with hypermethylation), Myc-DDK tagged pCMV6-Entry, pCMV6-Entry/GFP, intermediate-methylated (>5% alleles with intermediate and pCMV6-Entry/DKK3 plasmid vectors (Origene) were methylation), or unmethylated (no methylation detected). used for transient and stable expression. Transient transfec- tion was carried out in Opti-MEM medium using Lipofec- DNA copy number analysis (CNA) by qRT-PCR tamine2000 according to manufacturer’s recommendations DNA from 27 ACC samples that passed specified test (ThermoFisher Scientific) in 6-well plates with starting quality criteria were analyzed in quadruplicate with densities of 50,000 cells/well for SW-13 and 80,000 cells/ TaqMan Copy Number Assay using a primer / probe pair well for NCI-H295R cells. Cells were transfected one day specific to target gene DKK3 or housekeeping gene after plating. Transfection medium was replaced with ap- RPPH1. Normal adrenal tissue was used for diploid (2n) propriate growth medium 6 h post-transfection, and cells reference. Copy numbers were predicted using CopyCaller were assayed for cell behaviors 24 h post-transfection. Total software v2.0 (ThermoFisher Scientific). TaqMan Copy cell numbers and viability were calculated by staining cells Number Assay used was Hs00228043_cn. Target gene with 0.4% Trypan Blue (ThermoFisher Scientific) and DKK3 located on Chr.11:11989984 on NCBI build 37. counting with hemocytometer (Hausser Scientific Co.). Ex- Housekeeping gene Ribonuclease P RNA Component H1, periments were performed in triplicate, and parallel RPPH1 located on Chr.14:20811565 on NCBI build 37. pCMV6-Entry/GFP transfections were used to determine transfection efficiency. Immunofluorescence (IF) detection of Stable Geneticin (G418)-resistant pCMV6-Entry, Five μM-thick FFPE sections were processed for im- pCMV6-Entry/GFP, and pCMV6-Entry/DKK3 transfected munofluorescence detection of DKK3 and β-catenin clones were selected in 800 μg/mL G418-containing growth proteins as described previously [31]. Goat anti-DKK3 medium (ThermoFisher Scientific). Multiple clones were polyclonal (SC14959; 1:100 dilution) or mouse anti-β ca- then pooled into populations to avoid expression variability tenin monoclonal (SC47778; 1:200 dilution) primary and selection bias between clones. Established populations Cheng et al. BMC Cancer (2017) 17:164 Page 4 of 13

designated SW-Neo (from pCMV6-Entry transfections) clonogenic growth assays, cells were seeded in 6-well and SW-DKK3 (expressing Myc-DDK/DKK3) were com- plates at low densities (5,000 cells/well) and allowed to pared to parental SW-13 cells to determine effects of con- grow 7 days in appropriate growth medium (SW-Neo and stitutive DKK3 expression on SW-13 cells’ malignant SW-DKK3) with medium change every 3 days. On day 7, properties. Constitutive DKK3 expression was confirmed cells were washed with PBS, fixed, and stained as above. via qRT-PCR using TaqMan primer/probe pairs (Thermo- Colonies with 12 ± 2 or 4 ± 2 cells were counted as separ- Fisher Scientific) and Western blotting using anti-DKK3 ate groups and averaged from 6 wells. Experiments were mAb (1:500; Abcam), anti-mouse-HRP (Santa Cruz repeated 3 times, and data from a representative experi- Biotechnologies, Inc.), Mini-PROTEAN TGX gel, PVDF ment is presented. blotting membrane (Bio-Rad), and enhanced chemilumin- escence (ECL) detection reagents (ThermoFisher Scientific) Statistical analysis as per manufacturer’s protocols. Unless specified, 100 μg Normal distribution of continuous variables was assessed protein was loaded per well of 4–10% SDS gels (Bio-Rad). using D’Agostino and Pearson omnibus tests. Normally Equal protein loading was confirmed by staining PVDF distributed variables were analyzed using 2-tailed t test; membranes with GelCode Blue Safe Protein stain (Thermo- Mann–Whitney U test was used for non-normally dis- Fisher Scientific) after chemiluminescence detection. tributed variables. For variables with greater than 2 dependent values, a 1-way analysis of variance and Flow cytometric analysis of cell cycle Kruskal-Wallis tests were used for normally and non- SW-13, SW-Neo, and SW-DKK3 cells were fixed in cold normally distributed populations, respectively. Matched 70% ethanol for 30 min at 4 °C, washed twice with PBS, continuous variables were compared using Pearson cor- treated with ribonuclease (100 μg/mL), and stained with relation. Survival data were assessed by Kaplan-Meier propidium iodide (PI; 50 μg/mL in PBS). Using bandpass methods, and differences were compared by Mantel-Cox filter 605 nm (for PI), forward and side scatter were test. Statistical analyses were performed using Prism 6 measured in a BD LSRII Flowcytometer. Pulse process- (GraphPad Software). ing was used to exclude cell doublets from the analysis. FlowJo software was used to analyze the best Gaussian Results distribution curve to each peak for the cell populations Reduced expression of DKK3 in adrenocortical carcinoma of G0-G1 and G2-M. Recent comprehensive genetic analyses identified WNT signaling as the most common target of genetic aberra- Cell invasion, migration, adhesion, and clonogenic growth tions in ACCs. To identify novel WNT targets, we com- assays pared the expression pattern of selected positive and To assess invasive proficiencies, 100,000 SW-13, SW-Neo, negative WNT regulators in 7 ACC samples using an ex- or SW-DKK3 cells were allowed to invade through Matri- panded WNT expression array. Among various differen- gel from upper chambers containing serum-free medium tially expressed WNT regulators, the expression of DKK3, to lower chambers containing 10% FBS medium in BD a negative WNT regulator and a putative tumor suppres- BioCoat Matrigel invasion chambers (BD Biosciences). sor in a wide variety of tumors, was found significantly re- After 24 h, Matrigel was removed, and invaded cells were duced in the majority (6/7) of the ACC samples tested fixed in 3.7% formaldehyde/PBS (10 min), stained with (Fig. 1a; Additional file 1: Figure S1). Further, compared to 0.05% crystal violet (30 min), and counted at 100X magni- the robust expression pattern in adrenal cortex, DKK3 fication with light microscope. Matrigel invasion assay protein expression was found to be nearly absent in ACCs was performed twice in triplicate chambers. In migration by indirect immunofluorescence analysis (Fig. 1b; a&h). assays, 100,000 cells were allowed to migrate through DKK3 was observed to be expressed in the zona fascicu- 8 μM-pore size modified Boyden Chambers (BD Biosci- lata and zona reticularis (data not shown) in normal ad- ences) from upper chambers containing serum-free renal cortex, though to a lesser extent going inward from medium to lower chamber with 10% FBS medium. After 4 the zona glomerulosa (Fig. 1b). In contrast to the near ab- or 8 h, cells that migrated to lower side of the membrane sence of DKK3, β-catenin appeared to be over-expressed were fixed, stained, and counted as above. in ACC (Fig. 1b; h). Moreover, both robustly expressed Cell adhesion assays were carried out in 6-well plates. DKK3 and weakly expressed β-catenin proteins were One hundred thousand cells were seeded per well, allowed found predominantly in the cytoplasmic compartment of to grow overnight, washed with warm PBS, and incubated normal adrenal cortex (Fig. 1b; b-g), while increased β- with 0.5 mL of 0.25% Trypsin-EDTA for 1 min; Trypsin- catenin levels were found both in the cytoplasm and the EDTA was then removed, plates were tapped gently to re- nuclei of ACC cells (Fig. 1b; i-n). Due to the rarity of the move loosely attached cells, cells were washed with 10% disease and scarcity of fresh-frozen samples, an inter- FBS medium, fixed, stained and counted as above. For national patient cohort (n = 38) was assembled for DKK3 Cheng et al. BMC Cancer (2017) 17:164 Page 5 of 13

Fig. 1 Reduced DKK3 expression in ACC. a Reduced DKK3 gene expression in 7 tumor samples (T1–T7) compared to 3 histologically normal adrenal samples (N1–N3). T5a and T5b: RNA from two different areas of one tumor. Magnitude of gene expression relative to housekeeping gene panel shown below. b Immunofluorescence detection of DKK3 and β-catenin in normal adrenal cortex (a-g)andACC(h-n). Tissue sections treated with primary/secondary antibodies for DKK3 (FITC, green; c, j)orβ-catenin (TR, red; e, l), DAPI (blue for nuclear staining; b, i), or combinations of FITC/DAPI (d, k), TR/DAPI (f, m), or FITC/TR/DAPI (a, h, g, n). a and h: 100× magnification; b-g, i-n: 400× magnification; inset (g, n): 1000× magnification. c DKK3 gene expression (fold-change) in 37 ACC samples relative to average expression of 14 normal adrenal samples normalized to 1. d Average DKK3 expression (fold-change) in study cohort (n = 37) compared to average expression from 14 normal adrenal samples expression analysis (Table 1). Quantitative RT-PCR ana- patients with reduced DKK3 expression (p =0.19) lysis of 37 ACC samples confirmed reduced mRNA ex- (Additional file 1: Figure S3). pression in the majority (70%; 26/37) of ACC samples (Fig. 1c). The mean expression of DKK3 in 37 ACCs was DKK3 promoter methylation and gene copy number significantly decreased (p = 0.002) compared to mean alterations in ACC DKK3 expression in 14 normal adrenal tissue samples Promoter methylation has been identified as the principal (Fig. 1d). The high frequency of DKK3 silencing (70%) mechanism of DKK3 silencing in multiple tumor types observed in ACCs is very similar to that observed in other [34–39]. Moreover, we have previously shown potential malignancies including thyroid [32] and pancreatic involvement of global and gene-specific promoter methy- cancers [33]. lation changes in ACCs [40]. Using the EpiTect protocol To determine whether reduced DKK3 expression corre- [31], we analyzed methylation status of the DKK3 pro- lated with disease presentation and/or outcome, we ana- moter in 9 normal adrenal tissue and 29 ACC samples. lyzed statistical correlation to various patient characteristics Compared to the DKK3 promoter methylation status in (Table 1), including age, gender, tumor size, tumor weight, normal adrenal DNA, 4 ACC samples (14%) showed ENSAT stage, and hormone secretion phenotypes. Despite marked levels of hypermethylation, and 14 samples (48%) the limited cohort size (n = 38), reduced DKK3 expression showed intermediate-range methylation (Table 2). Twelve showed a non-significant trend (p = 0.062) towards female of 18 ACC samples with hyper- or intermediate promoter gender (Additional file 1: Figure S2). Kaplan-Meier survival methylation (67%) also showed significant reduction in analysis also did not reveal a significant effect on survival in DKK3 expression, concurring with the established role of Cheng et al. BMC Cancer (2017) 17:164 Page 6 of 13

Table 1 Summary of cohort demographics and patient Table 2 DKK3 mRNA expression, promoter methylation, and characteristics gene copy number alterations in adrenocortical carcinoma Characteristics Number of Cases Percentage Sample Gene Expression Promoter Methylation Gene Copy Number Total Number 38 NA 1 L IM 2 Gender 3 L HM 2 Male 14 35.8% 5 L HM 2 Female 24 63.2% 7 L UM 1 Age ± SD (y) 57.7 ± 13.2 NA 11 L UM 1 Cohort 13 L UM 2 Yale 10 26.3% 15 L UM ND Karolinska 25 65.8% 17 L UM 1 Düsseldorf 3 7.9% 19 L HM 1 Tumor Size (cm) 21 L IM 2 Mean ± SD 12.8 ± 4.4 NA 25 L IM 2 Range 5.5–21.0 NA 29 L IM 1 ENSAT 2008 Stage 31 L IM 2 I 0 0.0% 35 L ND ND II 18 47.4% 37 L ND ND III 11 28.9% 39 L ND ND IV 9 23.7% 41 L IM 2 Hormone Hypersecretion 43 L ND ND Aldosterone 1 2.6% 45 L UM ND Cortisol 9 23.7% 47 L IM 1 Androgen/DHEA 4 10.5% 53 L IM 1 Estrogen 1 2.6% 57 L HM 6 Multi-secretinga 5 13.2% 63 L UM 2 Non-functional 14 36.8% 65 L UM 3 No information available 4 10.5% 75 L ND ND Outcome 77 L ND ND Alive, no recurrence 11 28.9% 9NDIM 2 Alive, recurrence 5 13.2% 23 N UM 2 Death from disease 16 42.1% 33 N IM 2 Death from other causes 4 10.5% 49 N IM 2 Lost to follow-up 2 5.3% 71 N ND ND aTumors secreting two or more of the following hormones: aldosterone, 27 H IM 1 cortisol, testosterone, or DHEA y years, cm centimeter, SD standard deviation, ENSAT European Network for 51 H IM 3 DHEA NA the Study of Adrenal Tumors, dehydroepiandrosterone, 55 H IM 1 not applicable 59 H UM 2 promoter methylation in DKK3 silencing in other tumors 61 H UM 2 [36, 37]. Interestingly, 8/11 samples with non-methylated 69 H ND ND DKK3 promoters also showed comparable frequency of si- 73 H ND ND lencing (72%), suggesting alternate mechanisms for DKK3 Abbreviations: DKK3 Dickkopf-related protein 3, L low expression, N normal down-regulation in ACC. expression, H high expression, UM unmethylated, IM intermediate methylation, Recent genetic analyses of ACCs by us and others have HM hypermethylation, ND not determined shown significant copy number alterations in genes po- tentially involved in various signaling pathways [30]. To analyzed copy number variations using the TaqMan copy determine if gene copy loss contributed to reduced ex- number assay. We found copy losses in 9 samples (33%) pression of DKK3 in this cohort of ACC samples, we and copy gains in 3 of 27 ACC samples tested (Table 2; Cheng et al. BMC Cancer (2017) 17:164 Page 7 of 13

Additional file 1: Figure S4). Seven of the 9 samples with unperturbed and modifiable WNT signaling pathway, copy loss (78%) showed marked reduction in DKK3 ex- whereas the adrenal hormone-producing NCI-H295R cells pression; 4 showed concurrent DKK3 promoter methyla- harbor CTNNB1 and axin1 mutations, resulting in consti- tion. Interestingly, one ACC sample (ID #57) with 6 tutive WNT activation [31, 41]. To test whether suppressing copies of the DKK3 gene also showed promoter hyper- endogenous DKK3 will influence malignant properties of methylation and reduced expression of DKK3. ACC cells, we used a transient siRNA-silencing method. Si- lencing of DKK3 expression in SW-13 (Fig. 2b, d) and NCI- DKK3 silencing reduces clonogenic growth and promotes H295R (Fig. 2c) cells with siRNA was confirmed by qRT- migration of ACC cells PCR (Fig. 2b-c) and Western blot (Fig. 2d). DKK3 silencing To test whether DKK3 plays a tumor suppressor role in did not result in significant loss of viability in either cell type ACC in vitro, we investigated the expression pattern and for the duration of study (48 h). Due to low baseline levels regulation of DKK3 in two ACC cell lines, SW-13 and NCI- of DKK3 in H295R (Fig. 2a), siRNA-mediated silencing has H295R. Western blot analysis showed modest expression of no detectable effect observable by Western (data not DKK3 in SW13 cells, while NCI-H295R cells showed low shown). Next, we examined whether silencing of DKK3 expression (Fig. 2a). Despite carrying TP53 gene mutations, modulates clonal growth or migratory potential of ACC non-hormone-secreting SW-13 cells maintain an cells. Partial silencing (40% suppression; Fig. 2c) of DKK3

Fig. 2 RNAi silencing of DKK3 in ACC cell lines and effects on cell behavior. a Western immunoblot detection of endogenous DKK3 in SW-13 (left)and NCI-H295R (right). b and c, Relative expression of DKK3 as determined by qRT-PCR in siRNA-treated SW-13 (2B) and NCI-H295R (2C) cells, normalized to expression in cells treated with scrambled siRNA for 24 h. d Western immunoblot detection of DKK3 in SW-13 cells treated with control (1), scrambled negative siRNA (2), 10 (3), 20 (4), and 40 nM (5) DKK3 siRNAs for 24 h followed by protein extraction 48 h post-transfection. e-h, NCI-H295R (e and f)or SW-13 (g and h) cells treated with Lipofectamine (Lipo), scrambled negative siRNA (S-ive), or DKK3 siRNA (DKK3) for 24 h, allowed to grow in clonogenic growth conditions (e and g), or allowed to migrate through modified Boyden chambers through growth factor concentration gradient for 12 (f)or4(h) hours. Clones with 12 ± 2 cells were fixed, stained, and counted with light microscope (e and g); cells that migrated to lower side of modified Boyden chamber membranes were fixed, stained, and counted (f and h) Cheng et al. BMC Cancer (2017) 17:164 Page 8 of 13

(Fig. 2a) did not appear to influence clonogenic growth or Interestingly, 52% of the clones were small (4 ± 2 cells) migratory potentials of NCI-H295R cells (Fig. 2e-f). It is and composed of larger, slow-growing, or growth-arrested conceivable that the constitutively active WNT signaling in cells (Fig. 3e; right). In SW-Neo, this fraction of small these cells may have conferred inherent resistance to DKK3 clones represented only 9% of the clones (p < 0.001), while signaling. On the other hand, DKK3 silencing in SW-13 the remaining 91% constituted large colonies comprised cells (75% suppression; Fig. 2b; lane 5 of Fig. 2d) signifi- of 12 ± 2 cells (Fig. 3e; left). cantly impaired the cells’ ability to form colonies in isolation Next, we assessed the effect of constitutive DKK3 (p = 0.001) (Fig. 2g) and promoted their motility behavior over-expression on migratory potential of SW-13 cells. (p = 0.001) (Fig. 2h; Additional file 1: Figure S7). These re- SW-DKK3 cells exhibited significantly decreased migra- sults suggest a potential role for DKK3 silencing in adrenal tory potential compared to parental SW13 and SW-Neo carcinogenesis, which could be overrun by gain-of-function cells (p < 0.001) (Fig. 3f). To test whether DKK3- WNT mutations. promoted reduction in SW-13 cells’ migratory potential has a potential in vivo implication, we performed an in Exogenous DKK3 promotes migration of SW-13 cells vitro invasion assay. As reported previously in other can- Reports suggesting distinct roles for endogenous and se- cer types [26], over-expression of DKK3 significantly im- creted DKK3s in cell behavior [17, 42] prompted us to test paired SW-13 cells’ ability to invade through the effect of exogenous DKK3 addition to ACC cells. Cells reconstituted matrix (p < 0.001) (Fig. 3g). grown in the presence of exogenous human recombinant DKK3 did not show a difference in their overall growth DKK3 promotes a more differentiated phenotype in ACC potentials (Additional file 1: Figure S5). However, migra- cells tory potential of SW-13 cells was found to be accentuated To test whether decreased invasive behavior of DKK3- with exogenous DKK3 (Fig. 3a). The exogenous DKK3 in over-expressing SW-DKK3 cells is due to signaling this instance appears to have a dominant effect over the changes that can potentially modulate cell spreading and motility-impeding effect of endogenous DKK3 (Fig. 3a & thereby migration kinetics, cell morphology was ob- f). NCI-H295R cells with constitutively active β-catenin served under light microscopy. SW-DKK3 cells appeared appeared to be resistant (Fig. 3a) to the exogenous DKK3- to be larger with an extensive spreading phenotype aided induced migration-promoting effects on SW-13 cells. by dysregulated cell edge attachments (Fig. 4a-c). While the parental SW-13 and SW-Neo cells displayed a sig- Constitutive over-expression of DKK3 stifles malignant nificantly higher number of filopodia in a planar orienta- behavior of ACC cells tion, SW-DKK3 cells displayed a significantly higher DKK3 is constitutively expressed and persistently present proportion of lobopodial extensions (p < 0.01) (Fig. 4a- during zonal differentiation of adrenal cortex [14]. To test d). To test whether the differential expression of cell whether constitutive over-expression of DKK3 promotes extensions alters cell attachment characteristics, we per- redifferentiation of ACC cells, we generated a stable popu- formed a cell-detachment assay. SW-DKK3 cells showed lation of SW-13 cells engineered to over-express DKK3. a significantly stronger attachment to substratum com- Since NCI-H295R cells exhibited no appreciable response pared to both SW-13 and SW-Neo cells (p < 0.01) to either endogenous or exogenous DKK3 (Figs. 2e-f and (Fig. 4e). Whether increased attachment strength to sub- 3a), we limited our attention to SW-13 cells. Expression of stratum or multidirectional polarity conferred by the ectopic DKK3 was confirmed (Fig. 3b), and SW-DKK3 multitude of lobopodial attachments acts independently cells were assessed for various malignant properties com- or in tandem towards reduced invasive behavior of SW- pared to parental SW-13 and control SW-Neo cells. DKK3 cells needs to be studied further. SW-DKK3 cells grew at a slower rate compared to both parental SW-13 and SW-Neo cells (Fig. 3c). The slow rate FOXO1 as a potential DKK3 target to effect of growth of SW-DKK3 cells was found to be caused by redifferentiation an increase in the percentage of cells accumulated in G1 Towards understanding the potential transcriptional modu- phase (47.5% SW-Neo compared to 56.3% SW-DKK3 lation of cell adhesion and motility by DKK3 over- cells) of the cell cycle (Additional file 1: Figure S6). Since expression, we compared global difference in the expression suppression of endogenous DKK3 expression resulted in pattern of 84 transcription factors using an expanded tran- reduced clonogenic growth and increased motility of SW- scription array. Relative expression of 3 transcription 13 cells, we compared clonal growth and migratory poten- factors, ID1, JUN, and FOXO1, consistently demonstrated tial of SW-DKK3 cells to that of SW-Neo cells, using >4-fold difference in expression between SW-DKK3 and parental SW-13 cells as reference. Compared to their SW-Neo/SW-13 cells (Additional file 1: Figure S8 A&B). vector-transfected controls, SW-DKK3 cells showed an Transcription factors ID1 and JUN have been shown to me- overall increase in clonal growth efficiency (Fig. 3d). diate a variety of phenotypic effects, including apoptosis via Cheng et al. BMC Cancer (2017) 17:164 Page 9 of 13

Fig. 3 ACC cells were either treated with exogenous recombinant DKK3 (a) or enforced to express Myc-DDK tagged DKK3 (b) and assayed for cell behaviors. a SW-13 N/D (left) or NCI-H295R (right) cells were untreated (SW N/D-, 295-) or treated (SW N/D+, 295+) with exogenous DKK3 for 24 h and allowed to migrate through modified Boyden chamber for 4 h. Cells migrating to lower surface were fixed, stained, and counted. b Western immunoblot detection of endogenous DKK3 and ectopically expressed DKK3 (Myc-DDK/DKK3) in vector control (lane 1), SW-DKK3 (lane 2), or Myc-DDK/GFP control (lane 3)cells.c SW-13, SW-Neo, and SW-DKK3 cells plated in 24-well plates (5000 cells/well) were grown 8 days. Quadruplicate wells from each cell type were trypsinized, incubated in 0.2% Trypan blue, and viable cells were counted using hemocytometer. Data shown represent one of three independent experiments. d and e, Five thousand SW-13 or SW-DKK3 cells plated in 6-well plates were allowed to grow 7 days; clones were fixed, stained, and enumerated into 2 classes of (a)12±2cells(filled light grey) and (b) 4 ± 2 cells (filled black). Majority of clones formed from SW-Neo cells were large (e; left), while SW-DKK3 cells produced a significant number of small colonies (4 ± 2) comprised of large cells (e; right). f One hundred thousand SW-13, SW-Neo, and SW-DKK3 cells were allowed to migrate through modified Boyden chamber for 4 h; cells that migrated to the lower side of the membrane were fixed, stained, and counted. g One hundred thousand SW-13, SW-Neo, and SW-DKK3 cells were allowed to invade through Matrigel in modified Boyden chambers for 24 h. Cells that invaded through Matrigel and migrated to the lower side of the membrane were fixed, stained, and counted

DKK3 signaling, in multiple cancers (44, 45). DKK3-stifled magnitude of relief in migratory inhibition was found to be invasive behavior independent of loss of viability observed more pronounced in SW-DKK3 cells (45% increase in mo- in SW-13 cells prompted us to investigate a potentially tility with 43% FOXO1 suppression) than in SW-Neo cells novel role for FOXO1 transcription factor in DKK3- (30% increase in motility with 66% FOXO1 suppression; promoted redifferentiation of ACCs. Increased expression Additional file 1: Figure S10). These results clearly suggest of FOXO1 in SW-DKK3 cells was confirmed by qRT-PCR a role for FOXO1 in mediating DKK3-promoted redifferen- (Additional file 1: Figure S8C). Using siRNA, we transiently tiation and/or anti-invasive signaling in SW-13 ACC cells. silenced FOXO1 expression in SW-DKK3 and control SW- Neo cells (Additional file 1: Figure S9A&B) and assessed Discussion the effect of silencing on cell motility. Irrespective of DKK3 DKK3 expression is down-regulated in many human expression (Fig. 3b), both cell types showed an increase in cancers, including that of the thyroid, lung, prostate, migratory potential upon FOXO1 silencing (Fig. 5). The colon, breast, and liver [32, 33, 36, 43], but its Cheng et al. BMC Cancer (2017) 17:164 Page 10 of 13

Fig. 4 Constitutive over-expression of DKK3 reorganizes cellular extensions and cell spreading. a-c SW-13 (a), SW-Neo (b), and SW-DKK3 (c)cellswere grown on glass cover-slips, fixed, stained, and photographed. SW-13 and SW-Neo cells show a predominance of filopodia (red arrowheads) around edges; SW-DKK3 shows more lobopodia (small green arcs), absence of lamellipodia (blue arc), and few filopodia around edges. While cells in a and b appear to be polarized with filopodia at leading edge and lamellipodia at lagging edge, SW-DKK3 cells (c) show evenly spread flat lobopodia with extensive spreading and absence of polarity. Photomicrographs are taken using light microscope at 400× magnification. d Average number of lamellipodia, filopodia, and lobopodia per cell calculated from manual counting of cell extensions. Twenty randomly taken (400× magnification) photomicrographs of SW-13, SW-Neo, and SW-DKK cells used for quantification. e One hundred thousand SW-13, SW-Neo, and SW-DKK3 cells/well of 6-well plates were allowed to grow overnight, detached at specified times, cells remaining attached were fixed, stained, and counted manually regulation in ACC is unclear. In this study, we uti- Epigenetic modifications, including promoter methyla- lized comprehensive genetic, epigenetic, and func- tion and chromatin condensation, have been proposed tional approaches to identify and characterize a as major DKK3 silencing mechanisms in a variety of tu- potential tumor suppressor role for DKK3 in adrenal mors [43]. This study also supports a role for promoter carcinogenesis. Our study showed a significant de- hypermethylation in DKK3 silencing in ACCs. Interest- crease in DKK3 expression in 70% (25/37) of ACCs, ingly, DKK3 expression was also significantly decreased strongly suggesting a tumor suppressor role for DKK3 in many samples with intermediate methylation (48%), in human adrenal tissue. Whether the observed silen- suggesting that even intermediate levels of methylation cing in malignant samples represents an earlier dedif- may be adequate to silence DKK3 expression. Whether ferentiation or a later malignancy-promoting event the DKK3 promoter methylation observed in this study needs to be determined. Despite the relatively small is a component of the global methylation changes ob- cohort size, this study did not find an association be- served in ACCs [9, 40] or a specific DKK3 gene-targeted tween DKK3 silencing and prognosis, unlike in gastric event needs to be clarified. A large proportion of the cancer [35]. Of note, the majority of this cohort of ACC study cohort with non-methylated promoters but ACCs was previously shown not to harbor mutations with reduced DKK3 expression led us to seek alternate in DKK3 or FOXO1 genes while <10% carried beta- mechanisms for DKK3 down-regulation in ACC. In light catenin mutations [24]. of recent findings that gene copy number variations may Cheng et al. BMC Cancer (2017) 17:164 Page 11 of 13

did not affect growth or viability of cells but resulted in reduced clonogenic growth and increased motility, con- sistent with a tumor suppressor role for DKK3 [31, 41]. In contrast, exogenous addition of DKK3 to SW-13 cells resulted in increased motility, suggesting distinct roles for intracellular and secreted DKK3s. This observation is consistent with recent suggestions that DKK3 potentially has distinct intracellular signaling partners independent of canonical WNT-β-catenin circuitry [26]. Overall, intracellular DKK3 appears to confer a more differenti- ated phenotype to SW-13 cells. Whether the observed DKK3-promoted more differentiated phenotype is through (a) reactivation of the proposed adrenocortical differentiation pathway [14], (b) blocking of malignancy signaling networks, or (c) activation of a novel rediffer- entiation pathway needs to be clarified. Light microscopic analysis revealed drastic changes in organization of cell outgrowths on the edges of slow- Fig. 5 FOXO1 silencing releases DKK3-mediated block of cell migration. moving SW-DKK3 cells. While parental SW-13 and SW- Cells were treated with scrambled negative siRNA (Neg.) or FOXO1 siRNA Neo cells produced an overwhelming number of dynamic (Si-RNA) for 24 h, trypsinized, and allowed to migrate for 4 h through modified Boyden chamber. Migrated cells were fixed, stained, and filopodia that confer polarity and promote directional counted manually. Total number of SW-Neo cells treated with scrambled movement (41–43), SW-DKK3 cells showed predomin- siRNA normalized to 100%, and relative change in migration of SW-Neo antly lobopodia, indicative of multipolar spreading and and SW-DKK3 cells treated with FOXO1 siRNA is shown in overlapping hence arrested motility [44–46]. Although DKK3 has pre- line graph on left Y-axis. Relative FOXO1 expression in SW-Neo and viously been shown to influence migratory and invasive SW-DKK3 cells treated with FOXO1 siRNA, normalized to FOXO1 expression in SW-Neo cells treated with scrambled siRNA set at 100 phenotypes in multiple cancer cell types, an association of shownasbarsonrightY-axis cell surface modifications that can impact cell mobility has not been shown. The mechanism(s) that elicit the ob- served changes in cell extension repertoire need to be in- contribute to adrenocortical carcinogenesis [8, 24], we vestigated further. The association of the dedifferentiated analyzed a portion of our samples for DKK3 gene copy phenotype and loss of DKK3 expression in ACC, com- number variations. The majority of samples identified bined with the re-acquisition of a relatively more differen- with DKK3 copy loss also had significantly reduced tiated phenotype in SW-13 ACC cells overexpressing DKK3 expression. Only a handful of these samples had DKK3, suggest a global differentiation role for DKK3 in concurrent promoter methylation, indicating a possible adrenal cortex and the possibility that DKK3 could serve independent role for gene copy loss in causing DKK3 as a re-differentiation therapeutic target. down-regulation in ACC. One ACC sample with 6 cop- To explore potential pathways involved in eliciting the ies of DKK3 and a hypermethylated promoter had sig- observed DKK3-promoted redifferentiated phenotype of nificantly reduced expression of DKK3, suggesting that ACC cells, we compared the expression pattern of 84 hu- copy number variations may occur earlier in ACC onco- man transcription factors. Of the 3 transcription factors genesis than gene-specific methylation events. found to be over-expressed in SW-DKK3 cells (ID1, JUN Statistical correlation to patient characteristics and out- and FOXO1), FOXO1 secured our immediate attention for comes did not reveal any prognostic association of reduced 3 primary reasons: (1) FOXO1 is known to promote func- DKK3 expression in ACC patients, although reduced tional differentiation of myofibroblasts [47], (2) FOXO1 in- DKK3 expression was found to trend non-significantly to- hibits osteosarcoma malignancy via WNT inhibition [48], ward female gender. This study did not reveal a relationship and (3) FOXO1 transcription has been suggested in re- between DKK3 expression and aldosterone biosynthesis, as sponse to steroid hormones [49]. We hypothesized that reported earlier [25]. In addition, no significant correlation intracellular DKK3 promotes cellular differentiation signal- was observed in our tumor cohort between DKK3 expres- ing encompassing cellular spreading and stifled motility, at sion, metastasis, and tumor grade. least in part, via FOXO1 up-regulation. FOXO1 RNAi silen- We used functional approaches to characterize the ef- cing resulted in partial reversal of the motility suppression, fects of DKK3 on human ACC cells. Silencing of DKK3 suggesting that FOXO1 mayindeedplayaroleinDKK3- in SW13, a human ACC cell line with intact and induc- promoted redifferentiation of ACC cells. Based on the in- tile WNT signaling and endogenously expresses DKK3, verse relationship observed in ACC tissue between DKK3 Cheng et al. BMC Cancer (2017) 17:164 Page 12 of 13

and beta-catenin expression, it can be assumed that DKK3/ Competing interests FOXO1 regulation of malignant behavior of SW-13 cells is The authors declare that they have no competing interests. mediated through beta-catenin signaling. However, the pre- Consent for publication cise DKK3-FOXO1 signaling circuitry in the context of ad- Not applicable. renocortical differentiation needs to be investigated further. Ethics approval and consent to participate Written informed consent was obtained from patients prior to surgical Conclusions resection of adrenal tissue according to protocols approved by Institutional In conclusion, we demonstrate for the first time that Review Boards at (a) Yale University, New Haven, CT, USA, HIC#0812004538 DKK3 expression is frequently reduced in ACC, poten- (b) Heinrich Heine University Düsseldorf, Düsseldorf, Germany, and (c) Karolinska Institutet, Stockholm, Sweden. tially contributing to adrenal dedifferentiation and/or pro- gression of malignancy. Further, we identified FOXO1 as a Author details 1 downstream effector of DKK3 that may play a role in Department of Surgery & Yale Endocrine Neoplasia Laboratory, Yale University School of Medicine, New Haven, CT, USA. 2Department of blocking adrenocortical dedifferentiation. These results Pathology, Yale University School of Medicine, New Haven, CT, USA. suggest a potential for developing novel redifferentiation- 3Department of Oncology-Pathology, Karolinska Institutet, Karolinska 4 focused pharmaceuticals that could allow successful University Hospital, CCK, Stockholm, Sweden. Department of Surgery, Medical School, Heinrich Heine University, University Hospital Düsseldorf, treatment of ACCs when used in concert with existing Düsseldorf, Germany. 5Department of Nephrology, Medical School, Heinrich treatment regimens. Heine University, University Hospital Düsseldorf, Düsseldorf, Germany. 6Department of Surgery, Yale University School of Medicine, 333 Cedar Street, FMB130A, New Haven, CT 06520, USA. Additional file Received: 7 April 2016 Accepted: 22 February 2017 Additional file 1: Figure S1-S10. Supplementary Figures S1-S10. (PPTX 34931 kb) References 1. Lebastchi AH, Kunstman JW, Carling T. Adrenocortical carcinoma: current Abbreviations therapeutic state-of-the-art. J Oncol. 2012;2012:234726. ACC: Adrenocortical carcinoma; DKK3: Dickkopf-related protein 3; FBS: Fetal 2. Dackiw AP, Lee JE, Gagel RF, Evans DB. Adrenal cortical carcinoma. World J bovine serum; FF: Flash-frozen; FFPE: Formalin-fixed, paraffin-embedded; Surg. 2001;25:914–26. FITC: Fluorescein isothiocyanate; FOXO1: Forkhead Box Protein O1; 3. Ng L, Libertino JM. Adrenocortical carcinoma: diagnosis, evaluation and H&E: Hematoxylin & eosin; RPLP0: Ribosomal protein lateral stalk subunit P0; treatment. J Urol. 2003;169:5–11. RPPH1: Ribonuclease P RNA component H1; TR: Texas Red 4. Berthon A, Martinez A, Bertherat J, Val P. Wnt/beta-catenin signalling in adrenal physiology and tumour development. Mol Cell Endocrinol. 2012; – Acknowledgments 351:87 95. We would like to thank Drs. Irene Esposito and Henry Mayringer of 5. Lehmann T, Wrzesinski T. The molecular basis of adrenocortical cancer. – Department of Pathology, Heinrich Heine University, Düsseldorf, Germany for Cancer Genet. 2012;205:131 7. their contributions. 6. Simon DP, Hammer GD. Adrenocortical stem and progenitor cells: implications for adrenocortical carcinoma. Mol Cell Endocrinol. 2012;351:2–11. 7. Fonseca AL, et al. Gene expression and regulation in adrenocortical Funding tumorigenesis. Biology (Basel). 2012;2(1):26-39. This work was supported by the Damon Runyon Cancer Research 8. Assie G, Letouze E, Fassnacht M, Jouinot A, Luscap W, Barreau O, Omeiri H, Foundation (T.C.), the Ohse Research Foundation (T.C.B.), and a Medical Rodriguez S, Perlemoine K, Rene-Corail F, et al. Integrated genomic Student Research Fellowship from the NIH – NHLBI (J.C.). None of the characterization of adrenocortical carcinoma. Nat Genet. 2014;46:607–12. funding bodies played a role in data collection, analysis, or interpretation of 9. Fonseca AL, Healy J, Kunstman JW, Korah R, Carling T. Gene expression and data, the writing of the manuscript, or the decision to submit the manuscript regulation in adrenocortical tumorigenesis. Biology (Basel). 2012;2:26–39. for publication. 10. El Wakil A, Lalli E. The Wnt/beta-catenin pathway in adrenocortical development and cancer. Mol Cell Endocrinol. 2011;332:32–7. Availability of data and materials 11. Clevers H, Nusse R. Wnt/β-catenin signaling and disease. Cell. 2012;149: All data generated or analyzed during this study are included in this 1192–205. published article and its Additional file 1: Figure S1-S10. 12. Brisken C, Heineman A, Chavarria T, Elenbaas B, Tan J, Dey SK, McMahon JA, McMahon AP, Weinberg RA. Essential function of Wnt-4 in mammary gland Authors’ contributions development downstream of progesterone signaling. Genes Dev. 2000;14: JYC and TCB participated in all aspects of studies and contributed equally to 650–4. the manuscript. TCB conducted DKK3 expression, promoter methylation, CNV 13. Jordan BK, Shen JH, Olaso R, Ingraham HA, Vilain E. Wnt4 overexpression assays, Yale cohort clinical correlations, and statistical analyses, and drafted disrupts normal testicular vasculature and inhibits testosterone synthesis by pertinent manuscript sections. JYC performed cell culture, transfections, cell repressing steroidogenic factor 1/beta-catenin synergy. Proc Natl Acad Sci U growth, migration and invasion assays, Western blots, FACS analyses, drafted S A. 2003;100:10866–71. the rest of the manuscript, and was responsible for revisions. TDM and AS 14. Suwa T, Chen M, Hawks CL, Hornsby PJ. Zonal expression of dickkopf-3 and carried out transcription array analysis and FOXO1 siRNA studies. JMH components of the Wnt signalling pathways in the human adrenal cortex. J conducted WNT array analysis. AS, CCJ, and CL were responsible for KI patient Endocrinol. 2003;178:149–58. recruitment, disease characterization, and collection of nucleic acids, while WTK, 15. Schimmer BP, White PC. Minireview: steroidogenic factor 1: its roles in AK, and UIS were responsible for patient recruitment, disease characterization, differentiation, development, and disease. Mol Endocrinol. 2010;24:1322–37. and collection of nucleic acids in the Dusseldorf cohort. MLP performed 16. Heikkila M, Peltoketo H, Leppaluoto J, Ilves M, Vuolteenaho O, Vainio S. Wnt- histopathological analysis. RK performed light and fluorescence microscopy. TC 4 deficiency alters mouse adrenal cortex function, reducing aldosterone conceived the research design, secured funding for the study, and participated production. Endocrinology. 2002;143:4358–65. in experimental design and data analysis. All authors read, participated in the 17. Niehrs C. Function and biological roles of the Dickkopf family of Wnt modulators. critical revision of, and approved the final manuscript. Oncogene. 2006;25:7469–81. Cheng et al. BMC Cancer (2017) 17:164 Page 13 of 13

18. Bernichtein S, Petretto E, Jamieson S, Goel A, Aitman TJ, Mangion JM, 38. Lodygin D, Epanchintsev A, Menssen A, Diebold J, Hermeking H. Functional Huhtaniemi IT. Adrenal gland tumorigenesis after gonadectomy in mice is a epigenomics identifies genes frequently silenced in prostate cancer. Cancer complex genetic trait driven by epistatic loci. Endocrinology. 2008;149:651–61. Res. 2005;65:4218–27. 19. Howe LR, Subbaramaiah K, Chung WJ, Dannenberg AJ, Brown AM. 39. Chim CS, Pang R, Fung TK, Choi CL, Liang R. Epigenetic dysregulation of Transcriptional activation of cyclooxygenase-2 in Wnt-1-transformed mouse Wnt signaling pathway in multiple myeloma. Leukemia. 2007;21:2527–36. mammary epithelial cells. Cancer Res. 1999;59:1572–7. 40. Fonseca AL, Kugelberg J, Starker LF, Scholl U, Choi M, Hellman P, Akerstrom 20. Valencia A, Roman-Gomez J, Cervera J, Such E, Barragan E, Bolufer P, G, Westin G, Lifton RP, Bjorklund P, Carling T. Comprehensive DNA Moscardo F, Sanz GF, Sanz MA. Wnt signaling pathway is epigenetically methylation analysis of benign and malignant adrenocortical tumors. Genes regulated by methylation of Wnt antagonists in acute myeloid leukemia. Cancer. 2012;51:949–60. Leukemia. 2009;23:1658–66. 41. Durand J, Lampron A, Mazzuco TL, Chapman A, Bourdeau I. Characterization 21. Gaujoux S, Pinson S, Gimenez-Roqueplo AP, Amar L, Ragazzon B, Launay P, of differential gene expression in adrenocortical tumors harboring beta- Meatchi T, Libe R, Bertagna X, Audebourg A, et al. Inactivation of the APC catenin (CTNNB1) mutations. J Clin Endocrinol Metab. 2011;96:E1206–11. gene is constant in adrenocortical tumors from patients with familial 42. Papatriantafyllou M, Moldenhauer G, Ludwig J, Tafuri A, Garbi N, Hollmann adenomatous polyposis but not frequent in sporadic adrenocortical cancers. G, Kublbeck G, Klevenz A, Schmitt S, Pougialis G, et al. Dickkopf-3, an Clin Cancer Res. 2010;16:5133–41. immune modulator in peripheral CD8 T-cell tolerance. Proc Natl Acad Sci U 22. Mitsui Y, Yasumoto H, Nagami T, Hiraki M, Arichi N, Ishikawa N, Araki A, Maruyama S A. 2012;109:1631–6. R, Tanaka Y, Dahiya R, Shiina H. Extracellular activation of Wnt signaling through 43. Liang L, He H, Lv R, Zhang M, Huang H, An Z, Li S. Preliminary mechanism epigenetic dysregulation of Wnt inhibitory factor-1 (Wif-1) is associated with on the methylation modification of Dkk-1 and Dkk-3 in hepatocellular pathogenesis of adrenocortical tumor. Oncotarget. 2014;5:2198–207. carcinoma. Tumour Biol. 2015;36:1245–50. 23. Bertherat J, Groussin L, Sandrini F, Matyakhina L, Bei T, Stergiopoulos S, 44. Petrie RJ, Yamada KM. At the leading edge of three-dimensional cell migration. Papageorgiou T, Bourdeau I, Kirschner LS, Vincent-Dejean C, et al. Molecular J Cell Sci. 2012;125:5917–26. and functional analysis of PRKAR1A and its locus (17q22-24) in sporadic 45. Krause M, Gautreau A. Steering cell migration: lamellipodium dynamics and adrenocortical tumors: 17q losses, somatic mutations, and protein kinase A the regulation of directional persistence. Nat Rev Mol Cell Biol. 2014;15:577–90. expression and activity. Cancer Res. 2003;63:5308–19. 46. Petrie RJ, Gavara N, Chadwick RS, Yamada KM. Nonpolarized signaling 24. Juhlin CC, Goh G, Healy JM, Fonseca AL, Scholl UI, Stenman A, Kunstman JW, reveals two distinct modes of 3D cell migration. J Cell Biol. 2012;197:439–55. Brown TC, Overton JD, Mane SM, et al. Whole-exome sequencing characterizes 47. Vivar R, Humeres C, Munoz C, Boza P, Bolivar S, Tapia F, Lavandero S, the landscape of somatic mutations and copy number alterations in Chiong M, Diaz-Araya G. FoxO1 mediates TGF-beta1-dependent cardiac adrenocortical carcinoma. J Clin Endocrinol Metab. 2015;100:E493–502. myofibroblast differentiation. Biochim Biophys Acta. 1863;2016:128–38. 25. El Wakil A, Bandulik S, Guy N, Bendahhou S, Zennaro MC, Niehrs C, Mari B, 48. Guan H, Tan P, Xie L, Mi B, Fang Z, Li J, Yue J, Liao H, Li F. FOXO1 inhibits Warth R, Barhanin J, Lalli E. Dkk3 is a component of the genetic circuitry osteosarcoma oncogenesis via Wnt/beta-catenin pathway suppression. regulating aldosterone biosynthesis in the adrenal cortex. Hum Mol Genet. Oncogenesis. 2015;4:e166. 2012;21:4922–9. 49. QinW,PanJ,QinY,LeeDN,BaumanWA,CardozoC.Identificationoffunctional 26. Veeck J, Dahl E. Targeting the Wnt pathway in cancer: the emerging role of glucocorticoid response elements in the mouse FoxO1 promoter. Biochem – Dickkopf-3. Biochim Biophys Acta. 1825;2012:18–28. Biophys Res Commun. 2014;450:979 83. 27. Lee EJ, Jo M, Rho SB, Park K, Yoo YN, Park J, Chae M, Zhang W, Lee JH. Dkk3, downregulated in cervical cancer, functions as a negative regulator of beta-catenin. Int J Cancer. 2009;124:287–97. 28. Hoang BH, Kubo T, Healey JH, Yang R, Nathan SS, Kolb EA, Mazza B, Meyers PA, Gorlick R. Dickkopf 3 inhibits invasion and motility of Saos-2 osteosarcoma cells by modulating the Wnt-beta-catenin pathway. Cancer Res. 2004;64:2734–9. 29. Saeb-Parsy K, Veerakumarasivam A, Wallard MJ, Thorne N, Kawano Y, Murphy G, Neal DE, Mills IG, Kelly JD. MT1-MMP regulates urothelial cell invasion via transcriptional regulation of Dickkopf-3. Br J Cancer. 2008;99:663–9. 30. Brown TC, Juhlin CC, Healy JM, Stenman A, Rubinstein JC, Korah R, Carling T. DNA copy amplification and overexpression of SLC12A7 in adrenocortical carcinoma. Surgery. 2016;159:250–8. 31. Korah R, Healy JM, Kunstman JW, Fonseca AL, Ameri AH, Prasad ML, Carling T. Epigenetic silencing of RASSF1A deregulates cytoskeleton and promotes malignant behavior of adrenocortical carcinoma. Mol Cancer. 2013;12:87. 32. Yin DT, Wu W, Li M, Wang QE, Li H, Wang Y, Tang Y, Xing M. DKK3 is a potential in papillary thyroid carcinoma. Endocr Relat Cancer. 2013;20:507–14. 33. Hsieh SY, Hsieh PS, Chiu CT, Chen WY. Dickkopf-3/REIC functions as a suppressor gene of tumor growth. Oncogene. 2004;23:9183–9. 34. Kobayashi K, Ouchida M, Tsuji T, Hanafusa H, Miyazaki M, Namba M, Shimizu N, Shimizu K. Reduced expression of the REIC/Dkk-3 gene by promoter- hypermethylation in human tumor cells. Gene. 2002;282:151–8. Submit your next manuscript to BioMed Central 35. Yu J, Tao Q, Cheng YY, Lee KY, Ng SS, Cheung KF, Tian L, Rha SY, Neumann and we will help you at every step: U, Rocken C, et al. Promoter methylation of the Wnt/beta-catenin signaling antagonist Dkk-3 is associated with poor survival in gastric cancer. Cancer. • We accept pre-submission inquiries – 2009;115:49 60. • Our selector tool helps you to find the most relevant journal 36. Veeck J, Bektas N, Hartmann A, Kristiansen G, Heindrichs U, Knuchel R, Dahl • E. Wnt signalling in human breast cancer: expression of the putative Wnt We provide round the clock customer support inhibitor Dickkopf-3 (DKK3) is frequently suppressed by promoter • Convenient online submission hypermethylation in mammary tumours. Breast Cancer Res. 2008;10:R82. • Thorough peer review 37. Yang B, Du Z, Gao YT, Lou C, Zhang SG, Bai T, Wang YJ, Song WQ. Methylation of • Inclusion in PubMed and all major indexing services Dickkopf-3 as a prognostic factor in cirrhosis-related hepatocellular carcinoma. World J Gastroenterol. 2010;16:755–63. • Maximum visibility for your research

Submit your manuscript at www.biomedcentral.com/submit